Hướng dẫn cách sử dụng máy toàn đạc điện tử Topcon GTS 235

Topcon GTS-235 Total Station
Guide

--Jerry Davis, San Francisco State University IGISc

This guide is intended for using SFSU Institute for Geographic Information Science Topcon
GTS-235 total station (TS) to measure coordinates and recording the values in a data collector.
First, a checklist to make sure you have everything you need:
Topcon FC-100 Data
Collector with case.

USB cable*

Compass (if you want to set
up backsight to azimuth).

Prism + holder (in
accessories pack)

Serial cable
(data
collector
to TS) The TS uses the abbreviation NEZ (“Northing Easting Z”) for coordinates –
Some
background:
BC-27
this is the same as UTM, but remember that Northing comes before Easting,Topcon
in contrast
to battery
X
charger, output 9V, 1.8 A.*

AC/DC converter (for
coming before Y in standard Cartesian parlance.
charging data collector).
Power cord for battery
Topcon AD-9B. Output
use the
to measure coordinates, you must establish:
charger.
12V To
1000mA,
+ TS
center.*

1. The Occupied Point, where the TS is set up. This is either done by:
a. centering
the instrument over a known point (like a benchmark), and Raincover
entering the
data
Topcon
GTS-235
& cloth
b. resecting
to 2 or more known points using distance, thus requiring the rod-mounted target
(has removable
battery &
tribrach)
Range pole with integrated
prism
is held at those points.
level & standard

c. resecting to 3 or more known points using angles, like distant visiblebubble
landmarks.
Plummet
(plumbob)
5/8"-thread
brass mount for
2. A backsight to set the azimuth. It doesn’t use a compass, so initially it sets whichever
prism holder.
direction it’s looking to be 0° or north. If you have the Occupied Point established
in the TS,
Lens
cap
the backsight point will be used to accurately set the azimuth. If you use resection, the
Case (holds everything
backsight is not required.

A flat-top tripod, either
except data collector,
wooden or aluminum,
prism/rangepole assembly,
After
are established, readings are in the same coordinate system as thetripod,
pointsUSB
used.
Make
with these
5/8"-threaded
cable
&
suremounting

you are screw.
reading in meters if the coordinate system is UTM. If done right,
this isconverter
much more
AC/DC
for data
accurate than using a compass to establish the backsight direction.
collector)
* parts not needed in the
field (battery
chargersdirections step you through how to set up the instrument, thenNot
shown:
The following
how
to set tool
the kit with
and USB
cable)
jeweller's
screwdriver,
hex
occupied point and backsight, then how to get readings to record.
key, hook, 2 adjustment
pins and brush

Parts to carry:
1. Yellow case with total station, serial cable, plumbob, raincover, BC-27 battery charger
for total station, power cord, tool kit.
2. Tamrac accessories pack with prism, data collector, USB cable, AD-9B AC/DC, and
prism holder (shown in blue italics above and on next page)

3. Range pole with 5/8" mount
4. Flat-top tripod

Version 1/11/2009

1


Topcon Total Station Accessories Pack (Tamrac digital camera bag)
Prism
Topcon FC-100 Data Collector
Topcon AD-9B AC/DC Converter
Output DC12v 1000 mA center positive

space for compass, tools, etc.

in pockets:
• Prism rangepole mount (shown
here)
• USB cable
Other things you'll probably need:
• survey pins for instrument locations and
temporary benchmarks
• hammer
• survey tape (for instrument height, other
measurements)
• rebar (for monumenting if needed)

Instrument Setup
1. Make sure tripod legs are securely tightened and well set into ground surface, and that upper

mounting plate is as level as possible before mounting instrument. On sloping ground, have one
leg upslope. Carefully mount the instrument with the screw from the tripod. If you are going to
occupy a benchmark, position the instrument as near as possible over the benchmark; you
should use the plumbob to get it close (within a centimeter or so) – you'll be able to get it closer
after leveling using the laser plummet. You should remove the plumbob before continuing.

Packing the Total Station
The total station is a sensitive instrument. When packing the
total station in the case, make sure to unlock the horizontal
and vertical controls so they turn freely.

Version 1/11/2009

2


2. Identify the components you need to use:
a) The tribrach at the base, which contains
three leveling screws and a circular bubble
for coarse leveling. There's also a lock
which you can undo to release the tribrach,
but you won't need to do this.
b) Two identical LCD display units with:
•

Power

•

Menu

•

Escape

•

Enter

rough sighting
collimator
battery
Lens side of
telescope
Vertical
controls
Horizontal
controls

•

Star

•

NEZ coordinate mode

•

Horizontal distance mode

•

Angular measurement mode

Display

Laser
plummet

Tribrach
with leveling
screws

Horizontal & Vertical Controls
•
•

left-right and up-down
Function

vertical:
motion clamp
tangent screw

buttons used for menu choices.
c) A sighting telescope with eyepiece and lens
horizontal:
sides. Focus is the outermost brown circle, with
motion clamp

the cross-hair focus the tiny black ring around
tangent screw
the eyepiece. Above the telescope is a rough
sighting collimator with a triangle visible inside.
d) Horizontal and vertical control knobs: larger
motion clamps which are loosened for free movement and coarse adjustment, and
tightened for fine adjustments with the tangent screw, mounted on the motion clamp.
Located beside and above one of the display units and to the lower left of the telescope
lens. If it's on the eyepiece side, loosen the vertical motion clamp and rotate the
telescope 180° to get the lens on the clamp side – this is called the direct reading, and is
required for the data collector to work.
e) Plate bubble level, mounted above one of the displays.
f) A battery in the post to the left of the vertical control. The battery can be released with a
button on the top.
g) There are other parts like the laser plummet you shouldn't mess with, and signal and
power ports which you'll use later.
Check the battery level. You need to charge it before using it.

Version 1/11/2009

3


3. Coarse-level the instrument using the bubble level on the tribrach. Tip: Start by positioning
yourself in line with one of the adjustment knobs (C), with the other two knobs (A & B)
forming a perpendicular (see figure below). Then simultaneously adjust A & B to tilt the
instrument to the left or right; always turn them together either in or out from your
perspective – you'll be turning the knobs in opposite directions, but think in or out. Once
centered left to right, use the third knob C independently to center the bubble front to back.
You may have to do the left-right centering again, etc. You can also do this with one knob

(C) towards you.

AA

C

C

C

forward

B

AA

B

In: tilt left

Out: tilt right

(bubble right)

(bubble left)

AA

backward

B

4. Fine-level the instrument. Use the 30" (30 seconds of angle) plate level above one of the
LCD displays to do a fine adjustment. You'll use the same adjustment knobs, but you will
need to (loosen the horizontal motion clamp and) rotate the instrument parallel to the leftright axis, then parallel to the front-back axis, to do each of those adjustments. You'll notice
that the adjustment is very sensitive, requiring tiny movements, and the bubble moves slowly
to its stable reading – wait 5 seconds for it to adjust each time before you adjust any more.
Check the level in all four directions by rotating 90° each time. If your tripod is not mounted
on stable ground, you'll never get it leveled.
5. If occupying a benchmark, reposition over the benchmark using the laser plummet.
Note: One method that may make it easier to position over the benchmark is to remove the
instrument from the tribrach while positioning the tribrach over the point. Since it's locked,
you'll need to unlock it by first using the small screwdriver in the toolkit to undo the safety
on the lock, then unlocking and removing the instrument (carefully)
a) Power up the display. Wait a few seconds for the initial and contrast settings to display,
followed by the normal angle mode display (V and HR values displayed).
b) Press the star key to get to the star menu, and select the laser plummet with F4. The
laser plummet will then illuminate a visible red dot on the ground.
c) Loosen the tripod screw to position the dot on the benchmark; then tighten the screw.
d) Turn off the laser plummet to save power.
6. Extra-fine leveling using the tilt sensor. You can use the internal tilt sensor to get it leveled
even better. (If not already on, power the instrument up and wait for the normal angle mode
display.) Use the Star menu, then F2 to select the tilt sensor. Initially the X-On (F1) mode is
on, which you use to make tiny adjustments to the front-back tilt – the goal is to get it within
5" of 0. Then select XY-On (F2) to adjust the right-left. From here you can also make more
front-back adjustments. F4 lets you check the laser plummet if you want.
Version 1/11/2009

4

How to Collimate:
1. Point the telescope toward a bright
surface, like a blank white sheet of paper
held in front. Adjust the diopter ring
(closest to your eye) to focus the cross
hairs.
2. Aim the target roughly using the sighting
collimator, with the target at the peak of
the triangle in the sight.
3. Focus the target with the focusing knob.
from Topcon GTS 230 series manual

You may want to check for parallax by shifting your eye slightly when focused on a
distant object (> 200 m). If it moves, adjust the diopter and focus further to eliminate
parallax error.

What this Guide provides, and what are in other documents
This Guide:
Basic use for setting up the instrument with all related parts, for surveying points. May
be all you need, unless you need to go further or make adjustments. Assumes you'll use
the PDA to record data.
Topcon GTS 230 Series hardware manual
Assumes you're not using the PDA, and also provides information on the total station
parts, including checks for accuracy and adjustments.
Topsurv Manual
Covers the software that runs on the PDA. The GPS & Survey sections can be looked at
for more details on setting up the survey, etc. The GPS section isn't relevant so skip to
page 6-1 for the survey section. It covers some stuff this unit doesn't do, like scanning,
that you can ignore. The last section is on using the GTS for leveling.

Version 1/11/2009

5


Setting the instrument position: Option 1: Occupying a benchmark.
This is probably the most accurate method of re-surveying a site (See "Benchmarking and
Accuracy" section). If you've positioned the instrument over a benchmark, we just need three
settings established: (i) the height of the instrument; (ii) a backsight to establish direction in the
coordinate system being used; and (iii) the coordinates of the benchmark.
1. Connect the Data Collector (FC-100) using the serial cable from the 9-pin port on the bottom
right of the data collector to the circular 6-pin serial port on the DTS, and power it up.
2. Start TopSURV from the Windows desktop, and either select an existing job, or create a new
job. If you're starting a new data collection to later download, you'll probably want to create
a new job. Unless you know otherwise, use the default settings by pressing Finish when it
appears at the top.
3. Create a benchmark point by going to Edt > Points and Add. You can either give the point
a new number or go with the assigned (1). If desired, set a code; the pull-down will have no
existing codes, but you can create one simply by going to the entry and typing a code (like
"BENCHMARK" or "THALWEG" or something), then you'll have that code available later
as a pull-down selection. Enter the N (northing) E (easting) & Z (elevation) for the
benchmark. Ok out of this point, and close the Points dialog to get back to the main menu.
4. Use the Srv menu to get to Occ/BS Setup. Then select the Occ. Point you just created for
the benchmark. Measure the HI (height of the instrument) in the same units (probably
meters) of the coordinate system. Remember that you are surveying from the benchmark, so
the height of the instrument (HI) is important. All NEZ readings will be in reference to the
NEZ of the benchmark – the HI and rod height entries are used by the instrument or data
collector to make this work. If you later backsight to this point, the rod will be positioned on
the benchmark. You might want to see if the numbers make sense. Forgetting to enter the

correct HI or rod height will create a serious blunder in Z values
5. Consider what kind of backsight source you have. The possibilities are:
b. You know the true azimuth to some observable feature. You're only likely going to have
that if you use a compass (corrected to true azimuth with the known magnetic
declination.) In this case, you'll use enter this in the BS Azimuth field (the button toggles
with BS Point). Note that compasses are only accurate to about a degree.
c. You know the coordinates of distant features like mountaintops, tv broadcast towers, etc.
– too far away to occupy with the rod & reflector. In this case, you'll use BS Point with
Measure dist to BS unchecked (if doing multiple readings to check accuracy, you'll need
to set its Meas Type to HA/VA, which you'll find in its settings.)
d. You know the coordinates of nearby features that you can occupy with the rod &
reflector. You'll check Measure dist to BS, or for multiple readings set Meas Type to
HA/VA/SD.
6. Use either BS Azimuth or BS Point (more likely) to set the horizontal circle to true azimuths.
If you have multiple points, use these to verify accuracy.

Version 1/11/2009

6


BS Azimuth option: If you don't have a second
benchmark to sight to, but you do have the azimuth of
a visible feature to sight to (maybe established with a
compass), toggle to BS Azimuth and enter the azimuth
as DDD.MMSS (thus 206°34'08" is entered as
206.3408.) Use coordinate system azimuths, so if
you're using a compass for this, you'll need to know the
magnetic declination and the coordinate system
declination to enter the correct value. Then use the

button in the lower right to Measure the BS. Note the
BSCircle will display the reading from the TS, while
the azimuth will match the new setting.
Recommendation: Change the BS Circle to read the
same as the BS Azimuth by using its menu to Set to Az,
then transfer this to the total station with HC Set.
This allows you to read azimuths directly, and it may
make it easier to later re-establish the backsight.

Problem with Using a Compass for Backsight
Even if you adjust for magnetic declination, a 
compass will only be accurate to about a degree.  
The total station will then still be able to derive 
much more accurate angular readings, but they 
will all be off by the same amount as the compass 
reading was off.  Your map will be accurate 
geometrically and it should look ok since its 
orientation to north should be as close as the 
compass gets you.  But if you try to do a repeat 
survey later to measure change, using a compass 
to setup the backsight, those readings will all be 
off by the amount that new backsight was off.  
Thus the new survey will not align accurately with 
the old survey.   
 

Notes:
Alternative:  Use BS Point with fixed points 
• Remember that the instrument doesn't have a
compass, but instead has a highly accurate

horizontal circle, that reads from whichever direction it was pointing when you turn it on,
from internal memory. The backsight establishes a true direction.
• The memory feature allows it to remember its sighting direction while at one location
even if turned off and on. This is handy if something goes wrong and you need to start
the software over again.
• Recommendation: Change the BS Circle to read the same as the BS Azimuth – Use the
BS Circle menu to Set to Az, then transfer this to the total station with the HC Set. This
allows you to read azimuths directly from the instrument display, and it may make it
easier to later re-establish the backsight. If you haven't moved the instrument, you may
be able to enter the reading shown on the
instrument into BS Azimuth, then Meas to
make that equal to the BS Circle. But the
safest approach is to reshoot the backsight.
• You can sight to multiple backsights to
determine accuracy and possibly re-sight
individual backsights if you suspect an error.

Version 1/11/2009

7


BS Point option: If you have a points to sight to, use this option, even if you can't occupy it with
the rod. You must have the point entered in the point list,
so do this first if necessary using the method in step 3
Re-establishing an earlier survey's
above. Then measure to that point occupied with the rod,
coordinate framework
using its distance measurement, or to distant features that
If you're happy enough with the original

you can see but can't occupy (with Measure dist to BS
survey, and any compass-derived error
unchecked, or for multiple readings set the Meas Type to
isn't considered a problem, you can reHA/VA distance measurement). The instrument will derive
establish that framework later if you had
the azimuth vector from the two N-E coordinates.
established some fixed points – on
permanent stable features or
Where do you get the coordinates for the backsight points?
monumented with hammered-in rebar.
One source might be from a GPS survey, though you may
If from your instrument location (which
need to be concerned with limited accuracy. Submeter
you will need to have monumented for
accuracy units can work for this, but even their accuracy is
the repeat survey) you can sight to a
not as good as you can get with the total station. So think
fairly distant stable feature, you can
of distance as your friend. As seen in the Table 1 in the
record that azimuth to use for a
Benchmarking and Accuracy section, a GPS point accurate
backsight when you return. You'll still
to 20 cm at 20 m distance has an angular accuracy of 2063"
have the original compass backsight
(that's 2063 seconds of arc, or 0.57°), so this would be
error, but it is fixed so the surveys will
better than a compass reading. With a GPS point accurate
overlay accurately. So when you're
to 5 m, you would need to be at 500 m distance to see this
planning a repeat survey, even this

angular accuracy; this is beyond a reasonable sighting
method should only be used to establish
ranges. Greater distances reach finer accuracies, though
a coordinate system, and you'll need to
only survey-grade GPS readings to 0.01 m come close to
monument it so you can derive
the ~5" accuracy of the total station.
repeatable coordinates which you can
use for backsights.
Notes:
• Remember that the instrument doesn't have a compass, but instead has a highly accurate
horizontal circle, that reads from whichever direction it was pointing when you turn it on,
from internal memory. The backsight establishes a true direction.
• The memory feature allows it to remember its sighting direction while at one location even if
turned off and on. This is handy if something goes wrong and you need to start the software
over again.
• Recommendation: Change the BS Circle to read the
same as the BS Azimuth by using its menu to Set to
Az, then transfer this to the total station with HC
Set. This allows you to read azimuths directly from
the instrument display, and it may make it easier to
later re-establish the backsight. If you haven't
moved the instrument, you may be able to enter the
reading shown on the instrument into BS Azimuth,
then Meas to make that equal to the BS Circle. But
the safest approach is to reshoot the backsight.
• You can sight to multiple backsights to determine
accuracy and possibly re-sight individual backsights
if you suspect an error.

Version 1/11/2009

8


Setting the instrument position: Option 2: Resection to two or more benchmarks.
Though a bit less accurate, this option has the advantage of a somewhat easier setup since the TS
doesn't have to be positioned over the benchmark. You can however improve accuracy if you
have more than two benchmarks to sight to. Note that there are two types of resection: one that
only uses angles, requiring at least three benchmarks; the other using distance. We'll do the
latter. The first few steps are same as with option 1 above.
1. Connect the Data Collector (FC-100) using the serial cable, and power it up.
2. Start TopSURV from the Windows desktop, and either select an existing job, or create a new
job. If you're starting a new data collection to later download, you'll probably want to create
a new job. Unless you know otherwise, use the default settings by pressing Finish when it
appears at the top.
3. Use the Srv menu to get to Resection. You'll initially see a place to put an occupied point
for the instrument, but you don't have this so go to Next to get to your first benchmark to
sight to. If you want your derived instrument location to be used later as a benchmark, you
should enter an HI value, at the measured height of the instrument above the benchmark
you've created; then your saved position will be of the benchmark location below the
instrument, not the actual instrument. Select the point from the list, enter the rod height, and
press the Meas button once you've aimed the TS at the reflector. It should return with a
happy sound. Repeat for the next benchmark, etc. Check the map to see how the points are
arranged. Go to the Set tab to see how well the resection worked. You'll see values ('Sd' I
think is standard deviation) indicating how close your readings are. You can re-measure
readings to see if you can improve them; you can also remove a reading by going to
remeasure and exiting (not closing) back to the Set.
4. When you're happy with the results, press the Accept button to set the instrument location.
You'll be taken to the Store Point dialog, indicating that you want to store the instrument

position. Note that unless you had entered an HI value before, this position is that of the
instrument itself – this works fine, unless you need to re-occupy that point as a benchmark
later on.
Notes:
• Resecting to 3 or more points will produce an error check, and you might use this to note
problematic readings.
• If you are resecting, typically you are not at a permanent benchmark, so the HI (height of
the instrument) isn't needed. The NEZ for the instrument point is at the Z for the
instrument itself, not the ground beneath it. It's ok to enter an HI value, but make sure
you use it for both resection and foresights. You might want to check the numbers you
get to avoid blunders, especially in Z values.

Version 1/11/2009

9


Capturing points
At this point, we'll assume that your instrument position is known and it has a backsight
direction, from either of the above options. Now the instrument is ready to capture new points.
Use Srv > Observations to start capturing points. As with the point creation process, you have
a minimal data dictionary to use. (Perhaps you can make this more elaborate, but the basic
version works for me.) The Point is initially assigned a new number, but you can change this
and even mix letters or other characters. If you start the id with a letter, the program will
automatically use this as a prefix to increment (e.g., from 'A1' to 'A2', etc.). The Code is useful
for categorizing your operations, and you can create new types on the fly: existing code types
can be selected with a pull-down, or you can type in a new code, like "thalweg". One type of
point you may need is the next instrument position, where you should also hammer in a
benchmark. Note the rod height is easily seen and altered – you may need to change the rod
height in the middle of the survey to make it easier to see over brush. To survey a point, sight it

first, then press the Meas button. It will beep a happy sound if you get a good point, then will
prompt you for the HR – you don't need to change this; this just allows you to make a change if
you realize it's wrong. You will also be prompted for codes and such.
Moving the instrument
While you should try to maximize the number of points measured from a given instrument setup,
you will often find that you need to move the instrument to a new location. The most accurate
way to do this is to: (1) start from a benchmarked location, using a measured and entered HI
(height of the instrument); (2) survey in and establish the next position as a benchmark (either
with a temporary or permanent stake); (3) move the instrument to the next position, measuring
and entering its HI, and using the previous position as the backsight.
Exporting Data & Bringing it into a computer
To export a text file of points, use Job > Export … To File …
In the To File dialog, specify Data: Points, Format: Text then Next to go to a dialog for
specifying a name and location. (By default it goes into \CF Card\TPS TopSURV\IEFiles – but
check this anyway.) Then in text file format, specify comma or tab delimited, etc.
You can see your file in the FC-100 by opening it in MS WordPad (to keep TopSURV running,
press the windows button to get a start menu, and go to MS WordPad – when you exit it, you'll
be back in TopSURV.)
There are several ways of getting the data into another computer. The simplest is just to plug the
USB cable in, and set up Microsoft ActiveSync to use the connected device as a guest
(partnerships seem unnecessary). Using the location noted for export files (probably \CF Card
TPS TopSURV\IEFiles), explore for the text file you created. Open in Excel to convert the tabdelimited text. Remember that NEZ is essentially YXZ, not XYZ.
Importing Data from another computer
You can basically reverse the above process to import a file of coordinates. This is handy if you
have known points you want to import, maybe from a GIS file. You will probably want to
maintain the same fields. Use Job > Import to get to the right dialog.

Version 1/11/2009

10

Benchmarking and Accuracy
Accurate surveys require good control points, especially if a study area is going to be revisited to
determine change over time. This is commonly needed for riparian surveys, where floods can
move stream channel features. To attain centimeter accuracy, benchmarks such as 4' rebar must
be hammered into stable parts of the landscape. But how do we determine positions for these
benchmarks?
Local vs. Global Coordinate Systems. Traditional land survey methods typically use a local
coordinate system, and rely on benchmarks to maintain accuracy over time. These can be
converted into global coordinates such as the Geographic Coordinate System (GCS) of latitude
and longitude or cartesian coordinate systems based upon GCS such as UTM or state plane.
Critical to these is the datum system used, such as WGS84; survey manuals may refer to these
global coordinates as WGS84.
GPS provides global coordinates, but we must consider accuracy. If we have the time, situation
and equipment to derive centimeter-accuracy survey-grade GPS coordinates, we have a good
way to derive global coordinates for benchmarking. This is often not possible, so we're typically
using submeter-accuracy GPS for benchmarking. The best accuracy for these is attained through
post-processing. For the Trimble ProXH, we can probably get 20 cm or so accuracy, under good
conditions (pretty clear sky view, right time).
Position vs Angular Accuracy. First we need to consider the kinds of accuracy benchmarking
provides. There is both vertical (elevation) and horizontal (xy coordinates) accuracy, of course,
but there is also both positional and angular accuracy. Positional accuracy is fairly simple – if
you are off by some distance in x and y coordinates, that's your error.
Angular accuracy relates to position accuracy in two ways. (1) If a reading by the TS is off by a
given angle, the surveyed position can be off by the sine of the angle times the distance. With a
5-second TS like the GTS235, the relative error this creates is rarely a concern, though the
absolute angle depends on the backsight setting. (2) The positions of the from and to points used
for backsighting to set the HC will influence the accuracy of the HC, and thus of the positions
surveyed. Greater backsight distances are needed to reduce the effect of position errors on this

angular setting; a simple rule of thumb is
P
B= M
A
where: B = backsight distance (distance from the instrument to the backsight point)
P = backsight point accuracy
A = desired survey accuracy
M = maximum distance to surveyed point
So for a 50-m maximum survey distance, a 20-cm backsight point accuracy from a good GPS
point, to get a 1 cm accuracy from angular settings requires 50 * 0.2/0.01 = 1000 m backsight
distance. For 10 cm accuracy, that would be 100 m.

Version 1/11/2009

11


Note that this does not avoid the effect of position accuracy for the instrument; it only avoids
errors resulting from the angular HC setting. If the instrument position is off by 0.2 m, after
using a GPS to determine its position, then all of the points will be off by this much, in addition
to any other errors. This points to the need to use benchmarks for surveys needing repeating. If
a repeat survey depends on GPS to reestablish instrument position, the accuracy of the GPS
reading limits the accuracy of the survey.
Recommendations.
1. So for surveys that are going to be repeated, benchmarks are going to be needed, unless GPS
accuracy is sufficient for the changes expected.
2. To establish an correctly oriented coordinate system for mapping, GPS can be used for
benchmarking instrument and backsight position, as long as the distance is sufficient to
derive an accurate direction. Use the equation on the previous page to determine this, but
compare this with the accuracy of a compass (typically around 1°, or 3600 sec), which can be

improved upon with at least a 20 m backsight distance, given a 20-cm GPS accuracy:

Table 1. Angular accuracy in seconds from sightings to GPS points from 0.01 to 5 m accuracy

Position accuracy from 0.01 to 5 m 
dist 
(m) 
10 
20 
30 
40 
50 
60 
70 
80 
90 
100 
200 
300 
400 
500 
600 
700 
800 
900 
1000 

arc (m) 
from 1° 
0.17 

0.35 
0.52 
0.70 
0.87 
1.05 
1.22 
1.40 
1.57 
1.75 
3.49 
5.24 
6.98 
8.73 
10.47 
12.22 
13.96 
15.71 
17.45 

Version 1/11/2009

0.01 
206 
103 
69 
52 
41 
34 
29 
26 

23 
21 
10 
7 
5 
4 
3 
3 
3 
2 
2 

0.1 
2063 
1031 
688 
516 
413 
344 
295 
258 
229 
206 
103 
69 
52 
41 
34 
29 
26 

23 
21 

0.2 
4126 
2063 
1375 
1031 
825 
688 
589 
516 
458 
413 
206 
138 
103 
83 
69 
59 
52 
46 
41 

0.3 
6189
3094
2063
1547
1238

1031
884 
773 
688 
619 
309 
206 
155 
124 
103 
88 
77 
69 
62 

0.4 
8253
4126
2750
2063
1650
1375
1179
1031
917 
825 
413 
275 
206 
165 

138 
118 
103 
92 
83 

0.5 
10318
5157
3438
2578
2063
1719
1473
1289
1146
1031
516 
344 
258 
206 
172 
147 
129 
115 
103 

0.6 
12383
6189 

4126 
3094 
2475 
2063 
1768 
1547 
1375 
1238 
619 
413 
309 
248 
206 
177 
155 
138 
124 

0.7 
14450
7221
4813
3610
2888
2406
2063
1805
1604
1444
722 

481 
361 
289 
241 
206 
180 
160 
144 

0.8 
16519
8253
5501
4126
3300
2750
2357
2063
1833
1650
825 
550 
413 
330 
275 
236 
206 
183 
165 

0.9 
18589 
9285 
6189 
4641 
3713 
3094 
2652 
2321 
2063 
1856 
928 
619 
464 
371 
309 
265 
232 
206 
186 

1 
5 
m 
20661  108000 sec 
10318  52119  " 
6877  34539  " 
5157  25851  " 
4126  20661  " 
3438  17209  " 

2947  14746  " 
2578  12900  " 
2292  11465  " 
2063  10318  " 
1031  5157 
" 
688  3438 
" 
516  2578 
" 
413  2063 
" 
344  1719 
" 
295  1473 
" 
258  1289 
" 
229  1146 
" 
206  1031 
" 

12